Mean-variance analysis for supply chain management models

In the supply chain management literature under the stochastic environment, most of the proposed policies target at improving the supply chain's efficiency in terms of the expected cost reduction or the expected profit improvement. However, the performance measure with an expected value alone lacks precision when the corresponding variance is high. It also ignores the risk preferences of individual supply chain's decision makers. In order to provide a tailor-fit optimal decision-making policy for the decision maker, it is desirable to have a systematic and quantifiable measure for risk preference.; In light of all these, we study in this dissertation the application of the classical Mean-Variance Theory in finance for supply chain models. In mathematical finance, pioneered by the Nobel laureate Harry Markowitz in the 1950s, the Mean-Variance Theory has been an important theory for risk control in portfolio management. Under the Mean-Variance Theory, we can quantify the profit and risk in a portfolio investment by the expected return and variance of return, respectively. The Mean-Variance Theory has been demonstrated to be very applicable in practice. Based on the spirit of the Mean-Variance Theory, many optimal investment policies in finance are developed.; In this dissertation, using the idea of the Mean-Variance Theory, together with the Bayesian Decision Theory and the other optimization techniques, we study, analyze and build various supply chain management models, which include the inventory problems, the supply chain coordinating buyback contracts, and the optimal ordering policies with information updates.; This dissertation is divided into two parts and each part contains several chapters. Each chapter can be treated as a self-contained paper and the Mean-Variance Theory has been applied in each one of them. Throughout this dissertation, short example cases and numerical studies with computer simulations are included to illustrate the applicability of the models. From the studies in this dissertation, we can see that the classical Mean-Variance Theory can provide a systematic framework for the scientific studies of risk and uncertainty control in stochastic supply chain models in the information age. Moreover, the importance of risk control in supply chain management should not be neglected.